DevOps Classroom Series- SRE – 22/Oct/2019

Site Reliability Engineering – Intro

DevOps

1. Breaking Organizational silos Without Devops

With DevOps 2. Making frequent & smaller releases 3. Its ok to fail (Errors are normal) 4. Leverage Automation tools 5. Measure Everything

SRE

• Class SRE implments DevOps

How SRE implements DevOps

1. Breaking Organizational silos:
   • SRE in the team along with developers
   • Developers & SRE’s use same tools for deployments
2. Making frequent & smaller releases
   • Risk
3. Its ok to fail (Errors are normal)
   • We define Error Budgets
4. Leverage Automation tools
   • Toil
5. Measure Everything
   • Observability , Alerting etc

• Metrics:
  • SLI (Service Level Indicator)
    • At current point of time
    • eg. is service up or down
  • SLO (Service Level Objective)
    • At combine points of time
    • Is your service up for 99.9% time during last week/month
  • SLA (Service Level Agreement)
    • Service will up and running for 99.5% over a period of one month